By continuing to browse this site you agree to us using cookies as described in About Cookies
Wiley Online Library is migrating to a new platform powered by Atypon, the leading provider of scholarly publishing platforms. The new Wiley Online Library will be migrated over the weekend of February 24 & 25 and will be live on February 26, 2018. For more information, please visit our migration page: http://www.wileyactual.com/WOLMigration/
 The relationship between H3+ emissions in the Jovian sub-auroral region and auroral activity inside the auroral oval was examined using 3.53-μm image data from the IRTF Galileo support monitoring program of NASA. The sub-auroral emission region of H3+ was defined as a diffusive and weak emission region surrounding the auroral oval with a gradual decrease in intensity at lower latitudes beyond 6 RJ. The size of this region was found to vary approximately in proportion to the Jovian auroral activity. We discuss some possible sub-auroral emission mechanisms and suggest that the emission would be caused by precipitation from the inner magnetosphere associated with the higher latitude auroral activity, implying that the Jovian magnetospheric activity is tightly connected between the inner and middle magnetosphere.
 H3+ emissions in the Jovian middle and low latitudes have been detected by ground-based spectroscopic observations [e.g., Drossart et al., 1989; Lam et al., 1997; Miller et al., 1997]. Lam et al. , and Miller et al.  showed the global two-dimensional distribution of H3+ emissions, in which significant emissions were distributed in the middle and low latitudes. Miller et al.  argued that the emissions in the middle and low latitudes are caused by the transport of H3+ ions from the auroral region or by excitation due to particles precipitating from the inner magnetosphere. Infrared (IR) imaging observations of the Jovian aurora also indicated the faint emissions around the main auroral oval; these emissions are called the “polar collar” [Satoh et al., 1996] or the “Zone I aurora” [Satoh and Connerney, 1999a]. In this paper, we call the middle and low latitude emissions “sub-auroral emissions”.
Clarke et al.  identified weak FUV emissions equatorward of the main auroral oval by using HST observations. Clearer FUV images in the sub-auroral region are shown by Grodent et al. . They identified a secondary oval detached from the main oval in the equatorward. Further investigation of the FUV aurora [Prangé et al., 1998] revealed a low-latitude belt (LLB) distributed equatorward of the main oval and vanishing roughly near the footprint of Io's orbit.
 While the auroral emissions have been attributed to electron and ion precipitation from the middle and outer magnetosphere, it is unclear whether the sub-auroral emissions are related to auroral phenomena or to an atmospheric response such as gravity waves or solar radiation. In this paper, we present evidence that the sub-auroral H3+ emissions vary with the auroral activity.
2. Data and Data Processing
 The NASA Infrared Telescope Facility (IRTF) on Mauna Kea, Hawaii has been operating the Galileo support monitoring program since 1995. An NSFCAM infrared camera was mounted on the telescope, and daily standardized sets of observations of both Io and Jupiter have been provided under the management and operation of the University of Hawaii. We used two data-sets in the standard set of Jupiter observations: one for a filter with a central wavelength of 3.533-μm, corresponding to H3+ emissions, and the other for a 3.78-μm filter, which is sensitive to both NH3 clouds and H3+ emissions. The determination of the positioning of the Jovian disk on the 3.533-μm aurora frame was performed by using a 2-dimensional cross-correlation analysis between 3.78- and 3.553-μm auroras. Once the positioning is determined, we can decide the profile of Jupiter on the 3.533-μm aurora frame within an accuracy of about 1%. The field of view throughout the program was 37.9 × 37.9 arcsec. The image data-set for the H3+ emissions consists of four frame perspectives: northern background sky, Jovian northern hemisphere, southern background sky, and Jovian southern hemisphere frame. After subtracting the northern (southern) background sky frame from the northern (southern) hemisphere frame, we obtained the H3+ image data for each hemisphere.
 As the daily data-sets contain few standard-star observations, we used the averaged intensity of H3+ in the Jovian low latitudes as a reference to correct for terrestrial atmospheric extinction. Based on the report that the disk center brightness of H3+ has no correlation with the auroral emission [Satoh and Connerney, 1999b] and assuming that the lower latitude brightness of H3+ has no temporal variation, the average intensity of H3+ in the low latitudes in each frame is used to normalize the pixel data in the frame. The normalized value is called the “relative intensity”. We can compare any frame with any other frame using it without having to consider the daily variation in the earth's atmospheric conditions, although the intensity is not an absolute value.
 The analyses were carried out over the period from 1998 to 2000, during which we used 144 daily data-sets of good quality. Our analyses were restricted to the northern polar region because the declination of the earth (DE) was positive, meaning that the Jovian northern polar region was well observed during the period.
3. H3+ Emissions in Sub-auroral Region
 Examples of the H3+ emission distribution (“relative intensity” map) in the northern hemisphere are shown in Figure 1. The central meridian longitudes (CMLs) were selected to be close to each other to facilitate the comparison. The intensity is indicated both by a gray code and contour lines. The dotted red, green, and blue lines indicate ovals of 6, 12, and 30 RJ, which are the Jovicentric equatorial distances of the field line crossing calculated from the VIP4 plus current sheet model [Connerney et al., 1998]. A steep increase in the auroral intensity is apparent around the region where the relative intensity is around 9, and the familiar main-oval and polar-cap aurorae are almost included inside this region.
 Note in Figure 1a that there is a faint but widely distributed H3+ emission region surrounding the auroral region and expanding down to about 40–50° in the jovigraphic latitudes. We found that this faint brightness is not instrumental stray-light such as the scattering within a telescope, because the daily frame indicates a weak emission that always locates within the Jovian disk coinciding its border with the limb of Jupiter, nevertheless the daily frame has a different offset angle with respect to the limb of Jupiter. The latitudinal cross-sections of the H3+ emission for 25 October 1999 are shown in Figure 2. The emissions increased gradually from the middle latitude of about 6 RJ with a relative intensity of 3 to a higher latitude of around 10–12 RJ where the relative intensity was around 6–8, and then they steeply increased into the auroral region. The region of this gradual increase with a diffusive and structure-less nature can thus be discriminated from the auroral region. We call this region the “sub-auroral region”.
 The sub-auroral region seems to be phenomenologically a counterpart of the earth's diffuse auroral region connected to the central plasma sheet. However, it should be noted that the Jovian sub-auroral region sometimes extremely expands to lower latitudes beyond 6 RJ as can be seen in Figures 1a and 1b, which implies a different auroral process and/or the magnetosphere-ionosphere coupling system between Jupiter and the earth.
 The variability in the sub-auroral region can be seen by comparing Figures 1a, 1b, and 1c. The extent of the sub-auroral region is not constant and shows temporal variation. Moreover, the size of the sub-auroral region seems to be related to the auroral activity. In Figure 1c, the contracted sub-auroral region to higher latitudes appears to correspond to relatively less active H3+ emissions in the auroral region. Figure 1b may indicate the intermediate auroral activity and medium extent of the sub-auroral region. Figure 1a shows a large H3+ emission region corresponding to the third strongest aurora intensity during the analysis period.
4. Relationship Between Auroral Intensity and Sub-auroral Emissions
 To evaluate the relationship between the auroral intensity and the extent of the sub-auroral emissions, we need an index for the auroral activity. We accumulated the value of relative intensities that exceeds 8 in each observed image, and defined it as the “auroral intensity”. Two kinds of corrections were then performed to remove the geometrical and apparent modulations from the daily auroral intensities. The disk size modulation due to the variation in the orbital distance between the earth and Jupiter was corrected for by normalizing to a disk diameter of 40 arcsec. The apparent change in size of the auroral region due to the rotation of the tilted Jovian magnetic axis is also included in the auroral intensity. To remove this Jovian rotation effect, we used a method similar to that for IUE observations [e.g., Prangé et al., 2001]. The rotational effect was defined in this paper as a function of CML using a function fitting with a form of sin x + sin 2x. Then, all the observed auroral intensities were normalized using the fitted function. This normalized auroral intensity is defined as the “auroral activity index”.
Figure 3 shows the relationship between the auroral activity index and the sub-auroral emission area. The sub-auroral emission area is expressed by the pixel number of the area where the relative intensity is greater than 3, excluding the auroral region (>12 RJ) defined by IR aurora observations [Satoh and Connerney, 1999a]. We restricted the data used to estimate the sub-auroral region to the period when the Jovian northern pole faced the earth, corresponding to the CML range of 90°–240°. As shown in Figure 3, there was good correspondence between the auroral activity index and the extent of the sub-auroral region (correlation coefficient of 0.69). This means that the intensity of the auroral precipitation is closely related to the excitation of the middle and low latitude H3+ emissions and its extent.
 Because the emission mechanism of H3+ is thermal excitation, the observed intensity in the sub-auroral region could be related to any of three factors: H3+ ion density, environmental temperature, and H3+ production. Thus, when we consider the origin of sub-auroral H3+ emissions in relation to auroral activity, there are three possibilities. The first is H3+ transport to the sub-auroral zone [Miller et al., 1997]. If H3+ ions produced in the auroral atmosphere due to electron precipitation are transported by the neutral wind toward lower latitudes, the H3+ ions become abundant in the sub-auroral region, and H3+ emissions are emanated. The second is a hot neutral wind. If a hot neutral wind induced by the Joule heating in the polar ionosphere blows into lower latitudes, it would thermally excite the local H3+ ions. In these first and second mechanisms, the time for the transport to middle and low latitude is essential to consider the close relation between the auroral activity and the extent of the sub-auroral region. The Jovian Ionospheric Model (JIM) [Achilleos et al., 1998, 2001] developed after the theoretical discussion by Sommeria et al.  and Miller et al.  presented a simulation for the global three-dimensional dynamics of the Jovian thermosphere and ionosphere. Their results showed that H3+ ion distribution is enhanced inside the auroral belt, and the directions of both the ion and neutral motions are largely aligned with the auroral belt with a maximum speed of about 450 m s−1 (H3+ ion) and about 250 m s−1 (neutral). Even if the latitudinal velocity-component of these winds were 10% of the belt-aligned velocity, it takes 40000–20000 sec to travel 10 degrees of latitude. These long transport times are not plausible for interpreting the close relationship between the sub-auroral emission and the auroral activity. Moreover, the JIM simulation showed that the wind velocity is not enhanced even at higher altitudes, indicating that the dispersion into the sub-auroral region from the higher altitude (as suggested by Miller et al. ) would also be ineffective for the transport of H3+ ions. Therefore, the first two possibilities would be difficult to explain the relation between the sub-auroral emission and the auroral intensity.
 The third possibility is the direct production of H3+ ions in the sub-auroral region by precipitating electrons, which was first argued by Miller et al. . In this case, the precipitation in the inner magnetosphere (<10 RJ) must be concurrent with that in the middle and outer magnetosphere and must expand toward the planet as the auroral activity increases. The direct evidence for the particle precipitation into the sub-auroral region has not been disclosed. In the following, we discuss the possible precipitation and injection process based on previous research.
 Precipitation in the inner magnetosphere is considered to be due to the pitch angle scattering through wave-particle interaction because the induced H3+ emissions are mild and diffusive. Abel and Thorne  investigated the effect of the precipitation of the Jovian radiation belt electrons into the lower latitude atmosphere by using a weak diffusion scattering model and demonstrated that low latitude H3+ and X-ray emissions are partly caused by the precipitating energetic electrons. However, the L value of the central radiation belt (1.5–2 RJ) is too low compared with that of the sub-auroral emission region. Bhattacharya et al.  investigated the energy flux available for deposition into the Jovian upper atmosphere using in-situ measurements of energetic particles by Galileo. They showed that the enhanced population near the loss cone may provide precipitation energy through wave-particle interaction of an order of 40–100 ergs cm−2 s−1, which is enough to cause auroral emissions in the middle magnetosphere (10 < RJ < 25). On the other hand, their argument was rather negative for the emissions in the inner magnetosphere (<10 RJ) because the precipitation energy flux should decrease as the radial distance decreases. However, assuming that electrons are scattered at the strong diffusion limit, the measured electrons would deliver as much as 10 ergs cm−2 s−1 down to Io's orbit, which is sufficient to produce diffuse aurora [Prangé et al., 1998; Satoh and Connerney, 1999a]. We suppose that the Jovian energetic electron distribution has the potential to cause sub-auroral emissions in a wide area of the inner magnetosphere. On the other hand, the field aligned electron beam found by Frank and Paterson  may be a signature of magnetosphere-ionosphere coupling in this region. Here, we should note that the possibility remains that the precipitation of energetic ions due to the wave-particle interaction in the inner magnetosphere can also be the origin of the middle- and low-latitude aurora, as has been discussed by Thorne , Gehrels and Stone , and Mauk et al. .
 The injection of energetic particles into the inner magnetosphere can be seen in the Galileo observations. A quasi-periodic modulation of energetic particles, which reflects the global oscillation of the magnetosphere and the deep tail, has been identified by Woch et al.  and Krupp et al. . The large scale magnetospheric activity indicating a storm-like particle injection was reported by Mauk et al.  and Louarn et al. [1998, 2000]. Louarn  showed that high latitude auroral radio emissions took place concurrently with the multiple injections of energetic particles in the inner magnetosphere. Mauk et al.  also demonstrated that the inner magnetosphere dynamics are directly associated with the auroral emissions. These phenomena indicate that the magnetospheric disturbances are global and that the inner- and middle- magnetosphere are tightly connected in spite of the vast extent of the Jovian magnetosphere. The close relation between the auroral activity and the extent of the sub-auroral region derived from the present study would be the manifestation of the global disturbances in the Jovian magnetosphere.
 We would like to express our thanks to the staff of the IRTF for providing the data from the Galileo support monitoring program. This work was supported by a Grant-in aid for Scientific Research (124401299) from the Ministry of Education, Culture, Sports, Science and Technology, Japan.